ins-$74.0M · 13F 14+/11-×0.0 · short↓0.18
Current snapshotLAM RESEARCH CORP
AI Compute · Physical AI
AI Compute exposure in Physical AI.
- **LRCX** — Lam Research is identified via SEC EDGAR as a major supplier of deposition and etch tools for semiconductor manufacturing [SEC EDGAR LRCX CIK 00007
Catalysts
Sept-2026 quarter (Q1 FY2027) results Wednesday 2026-10-21 after market close, conference call 2:00pm PDT - nearest binary catalyst in the book (2 days out). Consensus EPS $2.18. Base to beat: revenue was $6.72B in the June-2026 quarter, and Lam issued guidance for the quarter ended 2026-09-27 alongside that release. WFE spend tied to AI/HBM capacity additions is the read-through; watch China revenue mix and services mix as the margin swing factors. [IR announcement 2026-09-30: https://investor.lamresearch.com/ ; events page https://investor.lamresearch.com/events ; June-quarter release 2026-07-29 https://newsroom.lamresearch.com/2026-07-29-Lam-Research-Corporation-Reports-Financial-Results-for-the-Quarter-Ended-June-28,-2026]
Invalidation
needs primary source validation — pending deep-dive research. Placeholder: (1) thesis chain breaks on primary-source evidence, (2) named revenue/customer concentration risk, (3) margin compression or competitive displacement
Latest change
· Coverage moved from — to Added to coverage.
Mapped funds and investors
Direct relationships supported by published evidence.
Reported holdings as-of
| Fund / investor | As of | Fund portfolio weight | Evidence |
|---|---|---|---|
| VanEck Semiconductor ETF SMH | 3.58% | Source | |
| iShares Semiconductor ETF SOXX | 3.24% | Source |
Disclosed investment history
No sourced relationships mapped yet.
Showing up to four relationships in each group. ETF weights describe the fund’s portfolio, not ownership of this company. Private investment history does not establish current ownership.
needs primary source validation — no X/Reddit signal yet
Lower-confidence snapshotTarget price: $376.62
Dated research targetMoat & position
- Evidence Grade
- Inferred
- Toll Road Tier
- Strong proxy
2026-10-21 · earnings
Deep research update
632 words · Updated 18 Oct 2026 · 8 sources
Research in development: this latest 632-word update is published for context while it is expanded toward Guide-level depth (1,000+ words).
LAM RESEARCH CORP is a US-listed company materially exposed to Advanced deposition / etch / CMP / thermal process tools.
Lam Research designs, manufactures, refurbishes, and services semiconductor wafer-fabrication equipment and related process technologies. In its own regulatory filings, Lam typically describes itself as a global supplier of wafer fabrication equipment and services to the semiconductor industry. The core economic model has two linked parts:
- System sales: Lam sells wafer-processing systems used to build logic, memory, foundry, and specialty devices. These systems are often large, multi-chamber platforms that perform etch, deposition, or clean steps.
- Customer support-related revenue: Lam sells spares, consumables, hardware upgrades, optimization services, training, and refurbished legacy systems. This is the recurring, higher-attach-rate portion of the model and is strategically important because semiconductor fabs cannot simply stop maintaining installed process equipment.
The system business is capital-intensive and cyclical; the customer-support business is more recurring and tends to grow with the installed base.
Lam Research’s 10-K typically reports revenue under two segments or groups: system revenue and customer support-related revenue.
The following figures are commonly cited from Lam’s fiscal 2024 Form 10-K, but because I could not resolve the exact EDGAR accession in this session, every numeric value is marked needs primary source validation:
- System revenue share: approximately 60–65% of total revenue —
needs primary source validation - Customer support-related revenue share: approximately 35–40% of total revenue —
needs primary source validation
Directionally, the model is: Lam sells a capital tool, then monetizes the installed base over the life of the tool through spares, services, upgrades, and legacy-system refurbishment. This is analogous to an “installed-base annuity” but the annuity is not contractual subscription revenue; it is purchase-order-driven service and parts demand.
Lam’s customer classes fall into three broad groups, which are discussed in company and competitor filings:
- Memory makers — DRAM and NAND producers. These buyers are highly cyclical and buy tools to add capacity or transition to higher-layer 3D NAND and advanced DRAM.
- Foundry / logic customers — companies that manufacture chips for external customers or for their own advanced logic products.
- Specialty / analog / power / MEMS / image-sensor customers — these customers use less advanced nodes but still require etch, deposition, and clean equipment.
Buyer archetypes include:
- Fab capex and procurement organizations: They evaluate total cost of ownership, throughput, yield, delivery, and service support.
- Process development and integration engineers: They evaluate process performance, repeatability, defect density, and integration with existing tools.
- Factory and yield teams: They evaluate uptime, service responsiveness, and spare-parts availability.
- Technical executives: They evaluate roadmap alignment, equipment supplier stability, and ability to support future nodes.
The company’s 10-K and competitor filings often discuss customer concentration and risk. The specific names likely appear in customer-concentration notes or risk factors. The following names are commonly cited in Lam and competitor filings as major semiconductor customers:
- Samsung — memory and foundry
- Micron — memory
- SK hynix — memory
- TSMC — foundry
- Intel — logic/foundry
- Kioxia — memory
- STMicroelectronics, Infineon, Analog Devices, NXP — specialty/analog, depending on the year
Do not treat this list as final. It must be verified against the latest 10-K exhibit. The customer names may appear in a risk-factor sentence such as “A small number of customers account for a substantial portion of our revenue,” followed by named examples or percentages.
- LRCX — Lam Research is identified via SEC EDGAR as a major supplier of deposition and etch tools for semiconductor manufacturing [SEC EDGAR LRCX CIK 00007
- - LRCX — Lam Research is identified via SEC EDGAR as a major supplier of deposition and etch tools for semiconductor manufacturing SEC EDGAR LRCX CIK 00007 [source
| Parameter | Value | Source | Tier | Measured/derived/assumed | Cross-verified |
|---|---|---|---|---|---|
| Total Lam Research fiscal 2024 revenue | ~$14.9B — needs primary source validation | Lam Research Form 10-K, FY2024; EDGAR accession not verified | Tier 1 | Measured | No — single-source pending verification |
| System revenue share | ~60–65% — needs primary source validation | Lam Research Form 10-K segment note; accession not verified | Tier 1 | Measured | No |
| Customer support-related revenue share | ~35–40% — needs primary source validation | Lam Research Form 10-K segment note; accession not verified | Tier 1 | Measured | No |
| Customer support-related revenue dollars | ~$5.9B — needs primary source validation | Lam Research Form 10-K segment note; accession not verified | Tier 1 | Measured | No |
| Installed base of chambers | ~80,000 — needs primary source validation | Lam Research 10-K/IR commentary; exact source not verified | Tier 2 | Measured/assumed | No |
| Gross margin | ~47% — needs primary source validation | Lam Research Form 10-K income statement; accession not verified | Tier 1 | Measured | No |
| Operating margin | ~29% — needs primary source validation | Lam Research Form 10-K income statement; accession not verified | Tier 1 | Measured | No |
| Top customer concentration | >30% of revenue — needs primary source validation | Lam Research Form 10-K risk factors; exact percentage not verified | Tier 1 | Measured | No |
| Annual customer-support revenue per installed chamber | $73,750 | Calculation from the two inputs above | Derived | Derived | No — formula-derived, not independently sourced |
All values in the table should be re-pulled from the actual EDGAR filing before publication. The table is not a final dataset.
_Pending primary-source company deep-dive._
Lam sells wafer-processing equipment and services. The primary product families, described qualitatively, are:
- Etch systems: plasma-etch tools that remove material selectively from patterned wafers. Lam’s etch portfolio includes conductor etch, dielectric etch, and bevel-etch/clean products.
- Deposition systems: tools that deposit thin films onto wafers. Deposition methods include chemical vapor deposition, atomic layer deposition, plasma-enhanced deposition, and electrochemical deposition.
- Clean systems: wafer-cleaning tools that remove particles, residues, and other contaminants between process steps.
- Support products: spares, software, process control services, training, and refurbished systems.
Within etch, Lam has historically sold product lines for high-aspect-ratio etch, conductor etch, dielectric etch, and bevel clean. Within deposition, Lam sells tools for both metal deposition and dielectric deposition. The precise product names and current models change from year to year; a 10-K product description should be consulted for the latest list.
The technical function of Lam’s equipment can be summarized as follows:
- Etch tools create a controlled plasma, use ion bombardment and reactive chemistry to remove material from the wafer, and rely on endpoint detection, chamber pressure, gas flow, and wafer temperature to maintain pattern fidelity. In advanced 3D NAND and leading-edge logic, etch is increasingly critical because smaller features and higher aspect ratios require more precise material removal.
- Deposition tools deposit thin films by chemical or physical means. ALD and ALE are increasingly important because they allow atomic-scale control of film thickness and etch depth.
- Clean tools remove residues and contaminants, which improves yield.
This is not commodity machining; it is process control under extreme conditions. The value is in the process chemistry, chamber design, software/recipe control, and integration with the customer’s process flow.
Substitutes for Lam’s products include:
- Competing etch/deposition tools from Applied Materials, Tokyo Electron, Hitachi High-Tech, and others. These can substitute for Lam tools in many process steps, though the cost of switching is high.
- Alternative process integration schemes. For example, a customer may choose a different device structure that requires fewer similar etch or deposition steps.
- Legacy tools or internally designed tools. Some very large integrated device manufacturers have historically explored internal tool development, but for high-volume manufacturing they primarily buy from established equipment suppliers.
- Lithography-centric approaches. ASML’s tools are not direct substitutes for etch/deposition, but changes in lithography can affect the number and type of etch/deposition steps.
Lam’s equipment is hard to replicate for several reasons:
- Process qualification: A fab must qualify a tool for a specific process. Qualification involves running test wafers, measuring yield, matching process results across chambers and tools, and integrating with adjacent steps. This can take months or years.
- Recipe and process libraries: Lam and customers co-develop process recipes over multiple technology nodes. The embedded know-how is often shared but not portable.
- Chamber and source design: Plasma source design, chamber materials, gas delivery, temperature control, and endpoint-detection software are not easily copied without violating intellectual property or taking years of R&D.
- Service and field-process support: Lam maintains field-process engineers and service technicians near major fabs.
- Scale and R&D spend: Lam’s R&D budget is large, and semiconductor equipment R&D is cumulative.
Lam can generally raise prices without losing share only in process steps where it is differentiated and where switching costs are high. For example:
- In high-aspect-ratio etch and certain deposition applications where Lam has market leadership, customers are less price-sensitive because tool performance affects yield and throughput.
- In more commoditized process steps, or where a competitor has achieved comparable performance, Lam faces more price pressure.
Exact pricing data is not disclosed in SEC filings. However, gross margin trends can proxy for pricing power. If Lam’s gross margin is rising while system revenue is stable or growing, that suggests pricing power or mix improvement; if gross margin is falling while revenue grows, that suggests price competition or an unfavorable mix.
_Pending primary-source company deep-dive._
The following conditions would weaken or invalidate a bullish or neutral view of LRCX:
- Durable collapse in memory fab capex. Lam has historically been sensitive to DRAM and NAND capital spending. If memory makers cut capacity additions for multiple years, system revenue and service demand would likely fall.
- Share losses in high-aspect-ratio etch or ALD/ALE deposition. If a competitor, especially Applied Materials or Tokyo Electron, demonstrates superior productivity or yield in Lam’s key process areas, Lam could lose qualified process steps. Share loss is difficult to reverse because requalification is costly.
- Gross margin compression without offsetting mix improvement. A decline in gross margin would suggest pricing competition or a less favorable product mix. This would directly challenge the pricing-power argument.
- Customer concentration realization. If one or two large customers reduce orders, Lam’s revenue can fall sharply. The 10-K may show significant customer concentration; a customer-specific capex cut is a key risk.
- Export-control or China-market disruption. Lam has significant exposure to China in certain years. If export controls tighten or China-based customers accelerate domestic tool substitution, Lam’s China revenue could fall.
- Installed-base service attach-rate decline. If fabs extend tool life, delay upgrades, or source third-party parts and service, the customer-support revenue per installed chamber would decline, reducing the recurring-revenue strength of the model.
- Technology-node transition delays. If leading-edge logic or 3D NAND transitions slip, demand for advanced etch and deposition can be pushed out.
- A competitor 10-K showing stronger gross margins or a faster-growing installed base in directly comparable etch/deposition segments. This would undermine Lam’s competitive-position claims.
In short, the thesis would be disproved if Lam loses process-step share in etch/deposition, if memory capex stays depressed, or if gross margins fall due to competition and export restrictions rather than improving later.
INTACT (exposure surfaced from technology map; no capital allocation)
Sources
8 sources preserved from the latest qualified research update.
- sec.govSEC EDGAR LRCX CIK 00007 Key evidence: LRCX — Lam Research is identified via SEC EDGAR as a major supplier of deposition and etch tools for semiconductor manufacturing SEC EDGAR LRCX CIK 00007 sourceOpen source ↗
- sec.govsec.govOpen source ↗
- sec.govsec.govOpen source ↗
- sec.govsec.govOpen source ↗
- sec.govsec.govOpen source ↗
- sec.govsec.govOpen source ↗
- sec.govsec.govOpen source ↗
- investor.lamresearch.cominvestor.lamresearch.comOpen source ↗